tarchon
First Post
Oh, this is one of those standards definition things. I was thinking it was a practical question. 
Well, I'm not sure why you would want to exclude non-letter characters (like if your users want to name something $^&*, should the software care?), but since you state the problem in the sense of exclusion, you must have some particular motivation that implies exclusion. Thus it would probably make sense if you stated the definition to cover all characters in the character set except the ones you want to exclude. You also have to decide what character set you will be using for such a definition to make sense of course.
If you really want to state it inclusively though while maintaining a broad range of character support, I think I would just make Unicode the char set and extract a list of all the codes that have the correct properties (probably 60,000 items or more).
If you don't use Unicode, you'll have to start looking at things in terms of typefaces or codepages too, and believe me, you don't want to go back to those dark, dark ages.
Well, I'm not sure why you would want to exclude non-letter characters (like if your users want to name something $^&*, should the software care?), but since you state the problem in the sense of exclusion, you must have some particular motivation that implies exclusion. Thus it would probably make sense if you stated the definition to cover all characters in the character set except the ones you want to exclude. You also have to decide what character set you will be using for such a definition to make sense of course.
If you really want to state it inclusively though while maintaining a broad range of character support, I think I would just make Unicode the char set and extract a list of all the codes that have the correct properties (probably 60,000 items or more).
If you don't use Unicode, you'll have to start looking at things in terms of typefaces or codepages too, and believe me, you don't want to go back to those dark, dark ages.